Asymmetric noncommutative torus has vanishing Einstein tensor.

Deeponjit Bose, Andrzej Sitarz · 2025

Read Asymmetric noncommutative torus has vanishing Einstein tensor. by Deeponjit Bose, Andrzej Sitarz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-15539.txt.

Introduction

Deeponjit Bose and Andrzej Sitarz compute geometric quantities for a noncommutative torus, a space where the order of multiplication matters, unlike ordinary geometry. They work with a spectral triple, a package of algebra, space, and a Dirac operator that encodes geometry in this setting. Their Dirac operator differs from a standard symmetric one by a partial rescaling that varies across the space.

Carrying out the calculation explicitly, they obtain the spectral metric, the torsion, and the Einstein tensor for this nontrivial example. Both the torsion and the Einstein tensor turn out to vanish identically. That flat outcome, despite the rescaling, makes the example a useful test case for theories of noncommutative gravity.

Catalog record

CREATORDeeponjit Bose, Andrzej Sitarz
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE92 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-15539

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Asymmetric noncommutative torus has vanishing Einstein tensor. Deeponjit Bose, Andrzej Sitarz Abstract We explicitly compute the spectral metric, torsion and Einstein tensors for a nontrivial spectral triple on a noncommutative torus, with the Dirac operator related to the fully equivariant Dirac by a partial conformal rescaling (as introduced in Dąbrowski and Sitarz 2015 ). The results show that the spectral triple has vanishing torsion and the Einstein tensor also identically vanishes. MSC Classification]58B34; 46L87 keywords Noncommutative Geometry; spectral triple, spectral geometry, Einstein tensor. pacs [ † † equal-contributors: These authors contributed equally to this work. 1 Introduction Spectral functionals, as introduced in Dąbrowski et al. 2023 are generalization of fundamental geometric tensors in the realm of noncommutative geometry and allow to compute the metric and Einstein tensors using spectral properties of the Dirac (or Dirac type operators) alone. Additionally, the spectral methods allow to determine a class of torsion-free Dirac operators without using the notion of linear connection. Though the construction has been tested on both classical and quantum geometries (including spin manifolds, Riemannian geometries with Hodge-Dirac operator, simple almost commutative geometries, conformally rescaled Dirac for the noncommutative tori and quantum SU(2) group) there are still many open problems. In Dąbrowski et al. 2023 we conjectured that suitably regular two-dimensional spectral triples will have the Einstein functional identically vanishing. In Dąbrowski and Sitarz 2015 we computed the scalar curvature for the so-called asymmetric noncommutative torus showing that the integrated scalar curvature is independent of the choice of the the Dirac operator, thus indicating this geometry as a nice testing ground to check the validity of the conjecture. In this paper we do prove that indeed the Einstein functional vanishes for this geometry, supporting our conjecture. Let us recall, that if \mathbb{T}^{2} is the classical torus with coordinates 0\!\leq\!x,y\!\leq\!2\pi , then we can equip it with the metric, dx^{2}+k^{-2}(x,y)dy^{2}, (1.1) where k is a strictly positive function. This is the mildest deformation of the flat metric. The scalar curvature of the torus with the metric ( 1.1 ) reads R=2k^{-1}\partial_{x}^{2}(k)-4k^{-2}(\partial_{x}(k))^{2}. (1.2) In the commutative case, such a metric is, of course, conformally equivalent to a flat metric on the torus. Using the spectral triples approach of Connes, where the natural object is the Dirac operator rather than the metric itself, we proposed a family of Dirac operators that provide a family of nontrivial geometries for the algebra of the noncommutative torus. In section 2, we briefly remind the construction of that spectral triple, the classical pseudodifferential calculus over the noncommutative torus, and compute the symbols of the Dirac operator and its inverse. In section 3 we recall the construction of spectral functionals and compute them for the spectral triple of the asymmetric torus. In section 4 we apply it and explicitly compute the relevant spectral functionals. 2 The spectral triple of the asymmetric torus. We start with the commutative case of Dirac operator on L^{2}(\mathbb{T}^{2},k^{-1}dx\,dy)\otimes\mathbb{C}^{2} for the metric ( 1.1 ): \tilde{D}_{k}=-i\sigma^{1}\left(\partial_{x}-\tfrac{1}{2}k^{-1}\partial_{x}(k)\right)-i\sigma^{2}\;k\,\partial_{y}, where \sigma^{1}=\left(\begin{array}[]{cc}0&\quad 1\\ 1&\quad 0\\ \end{array}\right),\quad\sigma^{2}=\left(\begin{array}[]{cc}0&\quad-i\\ i&\quad 0\\ \end{array}\right). (2.1) It is more convenient to work with the unitarily equivalent Dirac operator D_{k} acting on the dense domain of the Hilbert space H=L^{2}(T,dx\,dy)\otimes\mathbb{C}^{2} . D_{k}=-i\sigma^{1}\partial_{x}-i\sigma^{2}\left(k\,\partial_{y}+\tfrac{1}{2}\partial_{y}(k)\right). (2.2) The operator D_{k} is selfadjoint on the dense domain H . Next, we pass to the noncommutative torus \mathbb{T}_{\theta}^{2} , for which we refer to Connes 1994 for the details and following Dąbrowski and Sitarz 2015 we introduce the family of spectral triples. Definition 2.1 . Let \mathfrak{t} be the usual trace on \mathbb{T}_{\theta}^{2} and \mathcal{H}=L^{2}(\mathbb{T}_{\theta}^{2},\mathfrak{t})\otimes\mathbb{C}^{2} . Let \delta_{1},\delta_{2} be the usual self-adjoint derivations on \mathbb{T}_{\theta}^{2} and J be the standard real structure (Tomita-takesaki conjugation) on the coordinate algebra A(\mathbb{T}_{\theta}^{2}) . For a positive element k\in JA(\mathbb{T}_{\theta}^{2})J^{-1}\subseteq A^{\prime} , where A^{\prime} is the commutant of A , we define D_{k}=\sigma^{1}\delta_{1}+\sigma^{2}\left(k\,\delta_{2}+\tfrac{1}{2}\delta_{2}(k)\right), (2.3) obtaining a family of spectral triples (\mathbb{T}_{\theta}^{2},\mathcal{H},D_{k}) . The form of the operator D_{k} resembles the classical Dirac operator D ( 2.2 ), but since we take k in the algebra JA(\mathbb{T}_{\theta}^{2})J^{-1} which commutes with A(\mathbb{T}_{\theta}^{2}) , D_{k} has always bounded commutators with a\in A(\mathbb{T}_{\theta}^{2}) and it gives a spectral triple in the usual sense, representing the same K-homology class for each k . Moreover, D_{k} is a differential operator in the sense of Connes and Tretkoff 2011 and we can use the tools of the pseudodifferential calculus as used in Connes and Tretkoff 2011 ; Connes and Moscovici 2014 . Note that the bimodule of differential forms generated by a[D_{k},b] for all a,b\in A(T_{\theta}^{2}) is a free left module generated by \sigma_{1} and k\sigma_{2} . 3 Symbols of the Dirac For the details of the pseudodifferential calculus we refer the reader to Connes and Tretkoff 2011 ; Dąbrowski and Sitarz 2015 ; Dąbrowski et al. 2023 . To compute the spectral functionals, first we have to compute the symbols of the Dirac operator and its powers in the algebra of classical pseudodifferential symbols over the noncommutative torus. The square of D_{k} reads, \displaystyle D_{k}^{2} \displaystyle=\left((\delta_{1})^{2}+k^{2}(\delta_{2})^{2}\right), \displaystyle+\left(\tfrac{3}{2}k\delta_{2}(k)+\tfrac{1}{2}\delta_{2}(k)k+i\sigma^{3}\delta_{1}(k)\right)\delta_{2}, \displaystyle+\left(\tfrac{1}{4}(\delta_{2}(k))^{2}+\tfrac{1}{2}i\sigma^{3}\delta_{12}(k)+\tfrac{1}{2}k\delta_{22}(k)\right). and its symbol in the pseudodifferential calculus over A(\mathbb{T}_{\theta}^{2}) is, \rho(D_{k}^{2})=\mathfrak{a}_{2}+\mathfrak{a}_{1}+\mathfrak{a}_{0}, where, \displaystyle\mathfrak{a}_{2}= \displaystyle\left(\xi_{1}^{2}+k^{2}\xi_{2}^{2}\right), \displaystyle\mathfrak{a}_{1}= \displaystyle\left(\tfrac{3}{2}k\delta_{2}(k)+\tfrac{1}{2}\delta_{2}(k)k+i\sigma_{3}\delta_{1}(k)\right)\xi_{2}, \displaystyle\mathfrak{a}_{0}= \displaystyle\left(\tfrac{1}{4}(\delta_{2}(k))^{2}+\tfrac{1}{2}i\sigma_{3}\delta_{12}(k)+\tfrac{1}{2}k\delta_{22}(k)\right). Next, we compute the first three leading symbols of D_{k}^{-1} , \displaystyle\rho(D_{k}^{-1})=\mathfrak{b}_{-1}+\mathfrak{b}_{-2}+\mathfrak{b}_{-3}+\dots, (3.1) where b_{-n} is a symbol of D_{k}^{-1} homogeneous of degree (-n) in \xi . We have \mathfrak{b}_{-1}=(\sigma^{1}\xi_{1}+k\sigma^{2}\xi_{2})(\xi_{1}^{2}+k^{2}\xi_{2}^{2})^{-1}. The next homogeneous symbols we obtain from: \mathfrak{b}_{1}\mathfrak{b}_{-2}+\mathfrak{b}_{0}\mathfrak{b}_{-1}+\partial_{\xi^{j}}(\mathfrak{b}_{1})\delta_{j}(\mathfrak{b}_{-1})=0, which gives: \mathfrak{b}_{-2}=-\mathfrak{b}_{-1}\bigl(\mathfrak{b}_{0}\mathfrak{b}_{-1}+\partial_{\xi^{j}}(\mathfrak{b}_{1})\delta_{j}(\mathfrak{b}_{-1})\bigr), and that can be computed explicitly: \displaystyle\mathfrak{b}_{-2}= \displaystyle\sigma^{1}\biggl(kb_{0}\delta_{1}(k)b_{0}\xi_{2}^{2}-\frac{3}{2}kb_{0}\delta_{2}(k)b_{0}\xi_{1}\xi_{2}+kb_{0}^{2}\delta_{1}(k)b_{0}\xi_{1}^{2}\xi_{2}^{2} (3.2) \displaystyle-k^{2}b_{0}^{2}\delta_{1}(k)kb_{0}\xi_{2}^{4}+2k^{2}b_{0}^{2}\delta_{2}(k)kb_{0}\xi_{1}\xi_{2}^{3}-k^{3}b_{0}^{2}\delta_{1}(k)b_{0}\xi_{2}^{4} \displaystyle+2k^{3}b_{0}^{2}\delta_{2}(k)b_{0}\xi_{1}\xi_{2}^{3}-\frac{1}{2}b_{0}\delta_{2}(k)kb_{0}\xi_{1}\xi_{2}+b_{0}^{2}\delta_{1}(k)kb_{0}\xi_{1}^{2}\xi_{2}^{2}\biggr) \displaystyle+\sigma^{2}\biggl(-\frac{1}{2}kb_{0}\delta_{2}(k)kb_{0}\xi_{2}^{2}+2kb_{0}^{2}\delta_{1}(k)kb_{0}\xi_{1}\xi_{2}^{3}-kb_{0}^{2}\delta_{2}(k)kb_{0}\xi_{1}^{2}\xi_{2}^{2} \displaystyle-k^{2}b_{0}\delta_{2}(k)b_{0}\xi_{2}^{2}+2k^{2}b_{0}^{2}\delta_{1}(k)b_{0}\xi_{1}\xi_{2}^{3}-k^{2}b_{0}^{2}\delta_{2}(k)b_{0}\xi_{1}^{2}\xi_{2}^{2}+k^{3}b_{0}^{2}\delta_{2}(k)kb_{0}\xi_{2}^{4} \displaystyle+k^{4}b_{0}^{2}\delta_{2}(k)b_{0}\xi_{2}^{4}-b_{0}\delta_{1}(k)b_{0}\xi_{1}\xi_{2}+\frac{1}{2}b_{0}\delta_{2}(k)b_{0}\xi_{1}^{2}\biggr), where by b_{0} we denote b_{0}=(\xi_{1}^{2}+k^{2}\xi_{2}^{2})^{-1} . In the next step we compute the three leading symbols of D_{k}^{-2} , \displaystyle\rho(D_{k}^{-2})=\mathfrak{c}_{-2}+\mathfrak{c}_{-3}+\mathfrak{c}_{-4}+\ldots (3.3) where again, \mathfrak{c}_{-n} is a homogeneous symbol of D_{k}^{2} of degree (-n) in \xi . We obtain, \displaystyle\mathfrak{c}_{-2} \displaystyle=(\mathfrak{a}_{2})^{-1}=(\xi_{1}^{2}+k^{2}\xi_{2}^{2})^{-1}=b_{0}, (3.4a) \displaystyle\mathfrak{c}_{-3} \displaystyle=-\mathfrak{c}_{-2}\left(\mathfrak{a}_{1}\mathfrak{c}_{-2}+\sum_{i=1,2}\left(\partial_{\xi^{i}}(\mathfrak{a}_{2})\,(\delta_{i}(\mathfrak{c}_{-2})\right)\right) \displaystyle=\begin{aligned} &2\left(\xi_{2}\right)^{3}(\mathfrak{b}_{-2})^{2}k^{3}\delta_{2}(k)\mathfrak{b}_{-2}+2\left(\xi_{2}\right){}^{3}(\mathfrak{b}_{-2})^{2}k^{2}\delta_{2}(k)k\mathfrak{b}_{-2}-i\sigma^{3}\xi_{2}\mathfrak{b}_{-2}\,\delta_{1}(k)\,\mathfrak{b}_{-2}\\ &\hskip 5.69046pt+2\,\xi_{1}\left(\xi_{2}\right)^{2}(\mathfrak{b}_{-2})^{2}\left(k\delta_{1}(k)+\delta_{1}(k)k\right)\mathfrak{b}_{-2}-\frac{3}{2}\xi_{2}\mathfrak{b}_{-2}k\delta_{2}(k)\mathfrak{b}_{-2}\\ &\hskip 5.69046pt-\frac{1}{2}\xi_{2}\mathfrak{b}_{-2}\delta_{2}(k)k\mathfrak{b}_{-2},\end{aligned} (3.4b) \displaystyle\mathfrak{c}_{-4} \displaystyle=\begin{aligned} &-\mathfrak{c}_{2}\left(\mathfrak{a}_{1}\mathfrak{c}_{-3}+\mathfrak{a}_{0}\mathfrak{c_{-2}}+\sum_{j=1,2}\left(\partial_{\xi^{j}}(\mathfrak{a}_{1})\delta_{j}(\mathfrak{c}_{-2})+\partial_{\xi^{j}}(\mathfrak{a}_{2})\delta_{j}(\mathfrak{c}_{-3})\right)\right.\\ &\hskip 56.9055pt\left.+\frac{1}{2}\sum_{l,m=1,2}\left(\partial_{l}\partial_{m}(\mathfrak{a}_{2})\,\delta_{l}\delta_{m}(\mathfrak{b}_{-2})\right)\right).\end{aligned} (3.4c) 4 Spectral functionals In Dąbrowski et al. 2023 ; Dąbrowski et al. 2024 ; Dąbrowski et al. 2024 a family of functionals was introduced and studied for finitely summable spectral triples which have a Wodzicki-residue noncommutative trace Wodzicki 1987 (denoted as \operatorname{Wres} ) over the Connes-Moscovici pseudodifferential calculus. In the classical case of spectral triples over smooth functions of a Riemannian spin manifold and the Dirac operator being the usual spin Dirac differential operator these functional were demonstrated to reproduce the evaluation of geometric tensors on differential forms. The construction was shown to give the same tensors also for the Hodge-Dirac operator and was succesfully applied to some genuine noncommutative examples. We shall now compute explicitly these functional for the spectral triple defined in the section 2. 4.1 The metric functional Let us recall, that for any two one-forms in \Omega^{1}_{D}(A) the metric functional for a n -summable spectral triple over A is, \textnormal{{g}}_{D}(u,v)=\operatorname{Wres}\left(u\,v\,|D|^{-n}\right). (4.1) Proposition 4.1 . Let u_{i},v_{j}\in A(\mathbb{T}^{2}_{\theta}) , i,j=1,2 and u=u_{1}\sigma_{1}+ku_{2}\sigma_{2} , v=v_{1}\sigma_{1}+kv_{2}\sigma_{2} be the one-forms in \Omega^{1}(A(\mathbb{T}^{2}_{\theta})) . Then the metric functional for the spectral triple of the asymmetric noncommutative torus is: \textnormal{{g}}_{D_{k}}(u,v)=\tau\left(\frac{1}{k}u_{1}v_{1}+ku_{2}v_{2}\right) (4.2) Proof: We compute using results from section 3: \displaystyle\textnormal{{g}}_{D_{k}}(u,v) \displaystyle=\operatorname{Wres}\left(u\,v\,D_{k}^{-2}\right)=\int_{||\xi||=1}\mathfrak{t}\left(\operatorname{Tr}\left(\rho_{-2}\left(u\,v\,D_{k}^{-2}\right)\right)\right)d\xi (4.3) \displaystyle=\int_{||\xi||=1}\mathfrak{t}\biggl(\left(u_{1}v_{1}+k^{2}u_{2}v_{2}\right)(\xi_{1}^{2}+k^{2}\xi_{2}^{2})^{-1}\biggr) (4.4) \displaystyle=2\pi\tau\left(\frac{1}{k}u_{1}v_{1}+ku_{2}v_{2}\right). (4.5) where we have use that trace over the product \sigma_{1}\sigma_{2} vanishes and k commutes with u_{i},v_{j} . \qquad\Box 4.2 The torsion functional The torsion functional is a functional over three one-forms u,v,w , which is defined, for an n -dimensional

Read the complete text